Adversarial learning with multi-scale loss for skin lesion segmentation

Yuan Xue, Tao Xu, Sharon Xiaolei Huang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Inspired by classic Generative Adversarial Networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network with new activation function in the last layer as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale L1 loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. We show that such a SegAN framework is more effective in the segmentation task and more stable to train, and it outperforms current state-of-the-art segmentation methods in the ISBI International Skin Imaging Collaboration (ISIC) 2017 challenge, Part I Lesion Segmentation.

Original languageEnglish (US)
Title of host publication2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018
PublisherIEEE Computer Society
Pages859-863
Number of pages5
ISBN (Electronic)9781538636367
DOIs
StatePublished - May 23 2018
Event15th IEEE International Symposium on Biomedical Imaging, ISBI 2018 - Washington, United States
Duration: Apr 4 2018Apr 7 2018

Publication series

NameProceedings - International Symposium on Biomedical Imaging
Volume2018-April
ISSN (Print)1945-7928
ISSN (Electronic)1945-8452

Other

Other15th IEEE International Symposium on Biomedical Imaging, ISBI 2018
CountryUnited States
CityWashington
Period4/4/184/7/18

Fingerprint

Image segmentation
Skin
Pixels
Learning
Neural networks
Discriminators
Labeling
Labels
Chemical activation
Feedback
Imaging techniques

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Xue, Y., Xu, T., & Huang, S. X. (2018). Adversarial learning with multi-scale loss for skin lesion segmentation. In 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018 (pp. 859-863). (Proceedings - International Symposium on Biomedical Imaging; Vol. 2018-April). IEEE Computer Society. https://doi.org/10.1109/ISBI.2018.8363707
Xue, Yuan ; Xu, Tao ; Huang, Sharon Xiaolei. / Adversarial learning with multi-scale loss for skin lesion segmentation. 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. IEEE Computer Society, 2018. pp. 859-863 (Proceedings - International Symposium on Biomedical Imaging).
@inproceedings{6bd89c937d5b48d7bdbf89ad89ba3a2a,
title = "Adversarial learning with multi-scale loss for skin lesion segmentation",
abstract = "Inspired by classic Generative Adversarial Networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network with new activation function in the last layer as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale L1 loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. We show that such a SegAN framework is more effective in the segmentation task and more stable to train, and it outperforms current state-of-the-art segmentation methods in the ISBI International Skin Imaging Collaboration (ISIC) 2017 challenge, Part I Lesion Segmentation.",
author = "Yuan Xue and Tao Xu and Huang, {Sharon Xiaolei}",
year = "2018",
month = "5",
day = "23",
doi = "10.1109/ISBI.2018.8363707",
language = "English (US)",
series = "Proceedings - International Symposium on Biomedical Imaging",
publisher = "IEEE Computer Society",
pages = "859--863",
booktitle = "2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018",
address = "United States",

}

Xue, Y, Xu, T & Huang, SX 2018, Adversarial learning with multi-scale loss for skin lesion segmentation. in 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. Proceedings - International Symposium on Biomedical Imaging, vol. 2018-April, IEEE Computer Society, pp. 859-863, 15th IEEE International Symposium on Biomedical Imaging, ISBI 2018, Washington, United States, 4/4/18. https://doi.org/10.1109/ISBI.2018.8363707

Adversarial learning with multi-scale loss for skin lesion segmentation. / Xue, Yuan; Xu, Tao; Huang, Sharon Xiaolei.

2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. IEEE Computer Society, 2018. p. 859-863 (Proceedings - International Symposium on Biomedical Imaging; Vol. 2018-April).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Adversarial learning with multi-scale loss for skin lesion segmentation

AU - Xue, Yuan

AU - Xu, Tao

AU - Huang, Sharon Xiaolei

PY - 2018/5/23

Y1 - 2018/5/23

N2 - Inspired by classic Generative Adversarial Networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network with new activation function in the last layer as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale L1 loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. We show that such a SegAN framework is more effective in the segmentation task and more stable to train, and it outperforms current state-of-the-art segmentation methods in the ISBI International Skin Imaging Collaboration (ISIC) 2017 challenge, Part I Lesion Segmentation.

AB - Inspired by classic Generative Adversarial Networks (GAN), we propose a novel end-to-end adversarial neural network, called SegAN, for the task of medical image segmentation. Since image segmentation requires dense, pixel-level labeling, the single scalar real/fake output of a classic GAN's discriminator may be ineffective in producing stable and sufficient gradient feedback to the networks. Instead, we use a fully convolutional neural network with new activation function in the last layer as the segmentor to generate segmentation label maps, and propose a novel adversarial critic network with a multi-scale L1 loss function to force the critic and segmentor to learn both global and local features that capture long- and short-range spatial relationships between pixels. We show that such a SegAN framework is more effective in the segmentation task and more stable to train, and it outperforms current state-of-the-art segmentation methods in the ISBI International Skin Imaging Collaboration (ISIC) 2017 challenge, Part I Lesion Segmentation.

UR - http://www.scopus.com/inward/record.url?scp=85048127768&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048127768&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2018.8363707

DO - 10.1109/ISBI.2018.8363707

M3 - Conference contribution

AN - SCOPUS:85048127768

T3 - Proceedings - International Symposium on Biomedical Imaging

SP - 859

EP - 863

BT - 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018

PB - IEEE Computer Society

ER -

Xue Y, Xu T, Huang SX. Adversarial learning with multi-scale loss for skin lesion segmentation. In 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. IEEE Computer Society. 2018. p. 859-863. (Proceedings - International Symposium on Biomedical Imaging). https://doi.org/10.1109/ISBI.2018.8363707